is never truly going to be measurable and any response to that question is never going be any more than supposition.
The proportion of competence + honesty to incompetence + dishonesty is itself of secondary importance to the confluence of such incompentence + dishonesty and an unusually high potential for destructive consequences that positions of power permit. Whether or not "most" or a "few" politicians are corrupt or dishonest isn't as significant as the potential range of influence of the supposed incompetence + dishonesty.
In damage limitation of political influence, increasing the potential range of such influence is much worse than increasing or decreasing a constructed perception of the supposed incompetence or dishonesty.
It would presumably take thousands of ordinary people with poor impulse control rioting for an extended period to destroy a city.
It takes ONE incompetent and/or dishonest politican (supported by a personnel structure that supplies him or her with the means to do so) to acheive the same result.
Risk is the likelihood of something bad happening. Hazard is what you stand to lose if the bad things happen.
The risk of corruption being brought about by an individual politician or group of politicians is neither increased nor decreased by increasing or decreasing the hazard associated with their being corrupt.
Thousands more people have car accidents than plane accidents. Nobody stands in front of your car every morning performing a crash protocol lecture, and that's not because the risk is less but because the hazard is less.
Edit - had to remove some square brackets.